The times of swallowing events seen in ultrasound scans (US) were compared with the same measures as recorded using x-ray videofluorography (VFG). Twenty subjects, each on a protocol for a neuromuscular condition with possible dysphagia, completed three 10- cc water swallows that were recorded with US. Within the same hour as the US examination, each subject completed a VFG swallow of 7 cc with a liquid barium bolus. The following six events were compared from both techniques in sequence: (1) initial bolus movement (after being held steady on the front of the tongue), (2) the flattening of the back of the tongue, (3) start of hyoid excursion, (4) start and (5) finish of the hyoid in its maximum anterosuperior position, and (6) return of the hyoid to rest. Average swallow durations (from first bolus movement to the hyoid returning to rest) were consistent (VFG: 1.40 second, US: 1.23 second, P=.08). The timing of events was also consistent between both methods. Using the first bolus movement as a reference, the ratio of US times to VFG times were computed. US times were at least 81 percent of the VFG times in three measures (start and end of hyoid maximum, end of hyoid excursion) and at least 63 percent in two (start of tongue flattening, start of hyoid excursion) (100 percent would indicate equal measures). Additionally, durations among all six swallowing events were computed (15 different durations). The correlation coefficients between US and VFG durations were determined for each individual and for the average of all individuals. Nineteen of the 20 subjects had correlation coefficients above 0.81, with an average of 0.89. There was a very high correlation coefficient of 0.003 between the average US and VFG durations. The duration measurements completed in this study could not reliably predict pooling or dripping of the bolus into the valleculae. After allowing for variation in swallowing patterns, the results of US measurements are highly comparable to the VFG results of the same measures.